13  Assessing business performance

13.1 The story of Cisco

The following is a summary of a story from Rosenzweig (2014).

When Cisco Systems reached a market value of $100 billion in 1995, it had done so faster than any company in history. By March 2000, Cisco was the most valuable company in the world.

Through it’s rise, Cisco was frequently the subject of news articles documenting its rise and asking what was Cisco doing so well. Several factors were identified. One was the chief executive, John Chambers, who has learnt at IBM what happens if you don’t stay ahead of trends. He kept Cisco lean, humble and nimble.

Another factor was Cisco’s skill in acquiring companies. Cisco had great ability to select the best companies to buy, acquiring small companies that could immediately add to Cisco’s offerings rather than speculative or unrelated acquisitions. Cisco was also integrating the companies well.

Cisco’s peak in March 2000 coincided with the peak of the dot-com bubble. Despite a small correction, it held value through to September, continuing to garner articles from publications such as Fortune praising its CEO and operations.

But then the tech slide came. Cisco had lost more than half its value by the end of the year. By April 2001, Cisco had lost more than $400 billion in market capitalisation.

Then the negative articles began. In May 2001 Fortune wrote that Cisco’s prowess had not just been exaggerated, but had been false. Cisco had made its own mess through a cavalier approach to customers, signing long-term contracts at the wrong time, its acquisitions were poor and senior management had failed. Similar stories followed across the business press, the same publications that only a year ago were heaping praise on Cisco.

Over the next couple of years the tech sector slowly started to recover, and by 2003 Cisco was showing strong figures. In November of that year, John Chambers was on the cover of Business Week with the caption “Cisco’s comeback”. The article reviewed both what was going right in 2003, but also what had gone wrong in 2000. Cisco, one lauded for discipline and coordination, was out of control, with little coordinated planning. The previous assessment of disciplined acquisition was now seen to be a growth at any cost strategy through binge buying. What was once seen to be a focus on customers was now described as customer neglect.

This story provides two interesting observations. First, at any point in time the description of cause and effect seems plausible. But when looked at over time, it is clear somebody got the story wrong. Facts were assembled to tell the story of the moment - boom or collapse - rather than deep questions being asked about the causes of the performance. Cisco had not changed. Rather, a decline in performance led people to see the company differently.

Second, it highlights the difficulty in understanding company performance even as it occurs before our eyes. For all the attention Cisco achieved, it is not clear anyone understood the precise reasons for Cisco’s success.

13.2 The halo effect

Thorndike (1920) described his research into how superiors rated officers under their command. He found that the ratings of each person for traits such as intelligence, leadership and physical qualities were highly correlated. He suggested that the people giving these ratings were unable to analyse each specific trait for each person, so based the ratings on a general impression of the person as good or inferior. The ratings of each trait were suffused with a “halo” belonging to the individual as a whole.

Following Thorndike, this tendency to make inferences about specific traits based on a general impression is often labelled the “halo effect”.

The result of this difficulty is that people tend to allow good performance in the areas that they can measure (such as profits) to contaminate their assessment of other company attributes. They endow the company with a halo.

An experiment

An experimenter asked groups to project sales and earnings based on financial data. These groups were then given random feedback on their performance. Those with better feedback described their groups as cohesive, motivated and open to change, while those told they performed poorly said there was a lack of communication and poor motivation.

Staw (1975)

This tendency has implications for corporate decision making. When a researcher asks people to rate company attributes when they know the business outcome, those ratings are contaminated. If profits are up, people will assign positive attributes to that company. If times are bad, they will assign negative attributes. As for Cisco, all the factors responsible for a company’s rise might suddenly become the reasons for the fall, or be claimed to have never existed in the first place.

As an example, for many years Apple has had a clean sweep of all nine attributes in Fortune’s “World’s Most Admired Companies” poll. It is rated one in social responsibility, long-term investment value, innovation, people management and product quality. Is there not a single company in the world that is better than Apple on any of these nine? When asked nine questions, people don’t have nine different opinions. They just give their general impression nine times.

Understanding the halo effect is important when reading any book or article examining corporate performance. We need to ask whether the researchers designed the data collection such that their knowledge of the firm’s performance does not contaminate all that they do.

For further detail on this problem, see Rosenzweig (2007).